- Email: [email protected]
Differences in physical activity by gender, weight status and travel mode to school in Cypriot children
Available online at www.sciencedirect.com
Preventive Medicine 47 (2008) 107 – 111 www.elsevier.com/locate/ypmed
Differences in physical activity by gender, weight status and travel mode to school in Cypriot children Constantinos A. Loucaides a,⁎, Russell Jago b a
b
Research and Evaluation Unit, Cyprus Pedagogical Institute, Nicosia, Cyprus Department of Exercise, Nutrition and Health Sciences, University of Bristol, Bristol, U.K. Available online 14 February 2008
Abstract Objective. More information about children's physical activity during different periods of the day is needed. The purpose of this study was to describe children's physical activity during the segmented school day and examine potential differences during different periods of the day across gender, travel mode to school and weight status. Methods. School children (N = 247) wore pedometers for four consecutive school days and recorded their steps during the before school period, the 20-min school break, the whole school period, the after school period and for the whole day. Children also reported how they traveled to school. Data were collected in Cyprus, in January/February of 2007. Results. T-tests indicated that boys took significantly higher steps than girls throughout all the segments of the day ( p b 0.001) and children who walked to school exhibited higher step counts during the before school period ( p b 0.001), the after school period ( p b 0.01), and the whole day ( p b 0.01) in comparison to children who used motorized transport. A three-way ANOVA revealed significant two-way interactions between weight status and travel mode to school. Conclusion. Results suggest that there is a need to promote physical activity among children, especially among girls. Promoting active transport to school may further enhance the effectiveness of intervention programs. © 2008 Elsevier Inc. All rights reserved. Keywords: Pedometer; Children; Travel mode to school; Gender; Weight status
Introduction Participation in physical activity (PA) can promote children's musculoskeletal health, several components of cardiovascular health, reduce adiposity in overweight youth and improve blood pressure in mildly hypertensive adolescents (Strong et al., 2005). Promoting PA behavior among children in Cyprus could help reduce increased levels of overweight and obesity (Savva et al., 2002). Monitoring children's PA is an important first step for promoting this behavior. An objective and cost effective option for assessing children's PA is the pedometer. Electronic pedometers provide valid assessments of volume of PA performed by children (Trost, 2001; Bjornson and Belza, 2004). Identifying which periods of the day children are the least active could help target these ⁎ Corresponding author. 77 Larnaca Avenue, Aglanjia, 2102, Nicosia, Cyprus. E-mail address: [email protected] (C.A. Loucaides). 0091-7435/$ - see front matter © 2008 Elsevier Inc. All rights reserved. doi:10.1016/j.ypmed.2008.01.025
periods for promoting PA behavior. A recent study has quantified children's PA levels during distinct periods throughout the school day including the before school period, recess, lunch, physical education and after school activity (TudorLocke et al., 2006), but the study is limited by a small sample obtained from a single elementary school. No study has examined children's pedometer-determined PA during the segmented school day in relation to travel mode to school. Because active commuting to school is an important source of children's daily PA (Tudor-Locke et al., 2001; Jago and Baranowski, 2004), it would be interesting to examine whether children who walk to school differ in steps taken during specific periods of the day in comparison to those children who are driven to school. Previous studies using accelerometers in British (Cooper et al., 2003) and Danish children (Cooper et al., 2005), have shown that boys and girls who walked to school exhibited higher levels of overall PA in comparison to those with motorized transport. Providing data on travel mode to
108
C.A. Loucaides, R. Jago / Preventive Medicine 47 (2008) 107–111
school and PA from a different cultural perspective may help strengthen existing evidence. The aims of this study were to: 1) describe boys' and girls' pedometer-determined PA during the segmented school day, and 2) examine possible differences in PA levels during the segmented school day across gender, weight status and travel mode to school. Methods Participants Grade five (mean age 10.6 ± 0.3) and grade six (mean age 11.6 ± 0.3) Greek– Cypriot children from three urban public elementary schools in the coastal town of Lemesos, Cyprus were invited to participate in this study. In total, 296 children returned signed informed consent forms. Participation rates from the three schools were 75.2%, 73.9% and 66.4%. The Cyprus Pedagogical Institute and the Cyprus Research Promotion Foundation approved the protocol for this study.
Procedure PA was measured using an electronic pedometer (DW-200, Yamax Corporation, Tokyo, Japan). This pedometer has been shown to be valid and reliable among both children (Eston et al., 1998) and adults (Schneider et al., 2003). Prior to use, all pedometers were checked for accuracy and exhibited less than 5% error on a walk test and less than 1% error on a shake test (Tudor-Locke, 2002). Children were requested to wear the pedometers for four consecutive days. This monitoring frame has been shown to provide reliable estimates of children's PA (Trost et al., 2000). Data recording took place from Monday morning through to Thursday evening. In order to familiarize the children with the instrument, pedometers were given to the children the Friday of the previous week and were asked to fit the pedometer on Monday after awakening and dressing. Children were instructed to fit the pedometer at the waist and make sure that the cover was closed. They were also instructed to remove the pedometers only when bathing, swimming and sleeping. Instructions were given on how to reset the pedometer before reattaching it each morning. Children were given a recording card and were asked to record their steps during specific times of the day. The first recording was at 07:45 just before the school day started. This value represented children's before school activity. The second recording took place at 09:05, just before the 20-min break and the third recording at 09:25, just after the end of the break. The difference between these two values represented recess activity. The fourth recording took place at 13:05 just before the end of the school day. The difference between this value and the value recorded by children at 07:45, represented school activity. The fifth recording took place just before children went to bed. This value represented whole day activity and the difference between school activity and whole day activity represented after school activity. During school time, the recording cards were kept in the classrooms and class teachers supervised recording compliance. Two more 10-min breaks (10:45–10:55 and 12:15–12:25) are offered to
children during the school day. Children also noted how they traveled to school each morning (walk or car) and a further question also asked children to note their usual mode of travel to school (walk or car), in order to examine children's consistency in their reports. Height and weight were measured using a portable stadiometer and scale (SECA - Model 713) and Body Mass index (BMI) was calculated (kg/m2).
Statistical analyses Participants were included in the analyses if they had complete data for the 4 days of measurement. After screening the pedometer recording sheets, 38 children had incomplete data. Missing data were due to children being absent from school, children forgetting to fit the pedometer in the morning or losing the pedometer. Seven children were excluded from the analyses because of unreasonable step recordings (e.g., recording less step counts in the after break recording in comparison to the before break recording). Four participants were also excluded because of extremely high (N35,000) step counts per day (Rowe et al., 2004). Complete data were available for 247 (83.4%) children (124 boys and 123 girls). The percentage of boys and girls achieving the thresholds proposed by the President's Council on Physical Fitness and Sport (2001) (PCPFS) of 11,000 steps per day for girls and 13,000 steps per day for boys were calculated. BMI was computed as kg/m2. Children were classified as overweight and obese based on cut off points for BMI developed by the International Obesity Task Force (Cole et al., 2000). Means and standard deviations were computed for step counts across all five periods of the day. Independent-sample t-tests were conducted to examine potential differences in mean steps between boys and girls and between children who walked to school and those who were driven to school across the five periods. Finally, three-way ANOVAs were conducted to examine possible differences in steps in each of the five periods of the day across gender, travel mode to school and weight status. These were 2 (boys, girls) × 2 (car, walk) × 2 (normal weight, overweight) between subjects ANOVAs. Effect sizes (η2) were also calculated to examine the practical significance of the differences between group means (Cohen, 1988).
Results Mean BMI was 19.8 ± 3.8. The proportion of obese children in this sample was identical to national estimates (6.8 % versus 6.7%) but the proportion of overweight children in this sample was higher (29.6% versus 20.8%) (Savva et al., 2002). Sixty five of the children or 26.3 % reported traveling to school by foot and 179 or 72.5% reported traveling to school by car. About three quarters of boys (73.4%) exceeded the PCPFS threshold of 13,000 steps per day and 49.6% of girls exceeded the PCPFS threshold of 11,000 steps per day. Table 1 presents means, standard deviations and effect sizes of steps taken throughout the five different periods across boys and girls and the whole sample. Boys took significantly more
Table 1 Descriptives of steps and effect sizes during the segmented day across boys and girls and the whole sample in Cypriot children in January/February of 2007
Before school (awakening–07:45) 20-min Recess (09:05–09:25) School (07:45–13:05) After school (13:05–bed time) Whole day (awakening–bed time)
Boys (n = 124)
Girls (n = 123)
Mean (SD)
Mean (SD)
1227 (610) 1445 (585) 6030 (1873) 7964 (3353) 15221 (4468)
947 (423)⁎⁎⁎ 1039 (422)⁎⁎⁎ 4450 (1314)⁎⁎⁎ 5944 (2574)⁎⁎⁎ 11341 (3192)⁎⁎⁎
Effect Size (η2)a
All (n = 247) Mean (SD)
0.07 0.14 0.19 0.10 0.20
1088 (542) 1243 (549) 5243 (1799) 6958 (3151) 13289 (4338)
⁎⁎⁎Significant difference at the p b 0.001 level between boys and girls. Effect sizes of the significant differences in means between boys and girls. Values of 0.01, 0.06, and 0.14 can be interpreted as small, medium, and large effect sizes, respectively.
a
C.A. Loucaides, R. Jago / Preventive Medicine 47 (2008) 107–111 Table 2 Descriptives of steps and effect sizes during the segmented day across children who were driven and those who walked to school in Cypriot children in January/ February of 2007 Travel by car (n = 179) a
Walk (n = 65)
Mean (SD)
Mean (SD)
Before school 983 (487) 1355 (597)⁎⁎⁎ (awakening–07:45) 20-min Recess (09:05–09:25) 1244 (539) 1231 (584) School (07:45–13:05) 5215 (1861) 5329 (1653) After school (13:05–bed time) 6596 (2940) 8026 (3530)⁎⁎ Whole day 12795 (4290) 14710 (4278)⁎⁎ (awakening–bed time)
Effect Size (η2) b
0.08 – – 0.03 0.04
⁎⁎Significant difference at the p b 0.01 level between children who were driven and children who walked to school. ⁎⁎⁎Significant difference at the p b 0.001 level between children who were driven and children who walked to school. a Three children were inconsistent in their reports of travel mode to school and were thus excluded from the analysis. b Effect sizes of the significant differences in means between children who were driven and children who walked to school. Values of 0.01, 0.06, and 0.14 can be interpreted as small, medium, and large effect sizes, respectively.
steps than girls throughout all the periods of the day ( p b 0.001). Before school activity contributed to about 8.0% of both boys' and girls' daily PA, the 20-min recess period contributed to about 9.0% of daily PA, the rest of the school day about 31.0% of total daily activity and the after school activity about 52.0% of total daily activity. Table 2 presents means, standard deviations and effect sizes of steps taken throughout the five different periods across children who walked to school and children who were driven to school. Children who walked to school took significantly more steps than children who were driven to school during the before school ( p b 0.001), the after school ( p b 0.01), and the whole day period ( p b 0.01). Results of the three-way ANOVA (see Table 3) for the before school steps revealed a significant main effect for gender ( p b 0.001, η2 = 0.06), and a significant main effect for travel mode to school ( p b 0.001, η2 = 0.07). Boys took significantly more steps than girls and children who walked to school took significantly more steps than children who were driven to school. Analyses for the 20-min recess steps revealed a significant main effect for gender ( p b 0.001, η2 = 0.14). Identical results were revealed for the school steps ( p b 0.001, η2 = 0.16), with boys taking more steps than girls. The three-way ANOVA for the after school steps revealed a significant main effect for gender, ( p b 0.001, η2 = 0.10), a significant main effect for travel mode to school, ( p b 0.01, η2 = 0.04) and a significant travel mode by weight status interaction effect (see Table 3). Identical results were revealed for the whole day steps analysis. The ANOVA yielded a significant main effect for gender, ( p b 0.001, η2 = 0.19), a significant main effect for travel mode to school, ( p b 0.01, η2 = 0.04) and a significant travel mode by weight status interaction effect. Boys took more steps than girls and those who walked to school took more steps than those who were driven to school. The interactions showed that children who walked to school gained more steps than those who were driven during the after school
109
Table 3 Descriptives of steps during the segmented day across gender, travel mode and weight status and F-values of interaction effects in Cypriot children in January/ February of 2007 Boys Mean (SD)
Girls Mean (SD)
Normal weight
Overweight
Normal weight
Overweight
Before school Car 1148 (550)
1028 (589)
887 (390)
868 (396)
Walk 1472 (454)
1577 (932)
1158 (399)
1069 (567)
20-min Recess Car 1456 (596)
1394 (624)
1124 (453)
970 (376)
Walk 1495 (681)
1456 (326)
809 (196)
925 (523)
F-values (interaction effects)
F(1,226) = 1.59, p = 0.21 a F(1,226) = 0.09, p = 0.77 b F(1,226) = 0.23, p = 0.63 c F(1,226) = 0.86, p = 0.36 d
F(1,226) = 2.01, p = 0.16 a F(1,226) = 0.04, p = 0.85 b F(1,226) = 0.81, p = 0.37 c F(1,226) = 0.58, p = 0.45 d
School Car 6152 (2030) 5736 (2132) 4604 (1367) 4143 (1084) F(1,226) = 0.20, p = 0.66 a Walk 6074 (1477) 6008 (1687) 4327 (1430) 4155 (1095) F(1,226) = 0.02, p = 0.88 b F(1,226) = 0.39, p = 0.53 c F(1,226) = 0.00, p = 0.95 d After school Car 8212 (3433) 6372 (2648) 6207 (2438) 4791 (1977) F(1,226) = 0.95, p = 0.33 a Walk 8717 (3236) 9661 (3935) 5861 (2118) 7176 (4142) F(1,226) = 0.20, p = 0.66 b F(1,226) = 9.37, p = 0.01 c⁎⁎ F(1,226) = 0.00, p = 0.98 d Whole day Car 15512 (4769) 13136 (4319) 11699 (3117) 9802 (2686) F(1,226) = 1.24, p = 0.27 a Walk 16263 (3598) 17245 (4016) 11345 (2911) 12400 (4252) F(1,226) = 0.06, p = 0.82 b F(1,226) = 7.18, p = 0.01 c⁎⁎ F(1,226) = 0.03, p = 0.86 d ⁎⁎Significant at the p b 0.01 level (η2 = 0.04 for the after school steps and η2 = 0.03 for the whole day steps). a Gender by travel mode to school interaction effect. b Gender by weight status interaction effect. c Travel mode to school by weight status interaction effect. d Gender by travel mode to school by weight status interaction effect.
110
C.A. Loucaides, R. Jago / Preventive Medicine 47 (2008) 107–111
Fig. 1. Bar chart showing the interaction between travel mode to school and weight status during the whole day activity in Cypriot children in January/ February of 2007.
and the whole day steps, but this difference was only evident among overweight children ( p b 0.01). The interaction for the whole day steps is presented graphically in Fig. 1. Based on the regression equation by Beighle and Pangrazi (2006) suggesting that 5000 steps results in the accumulation of 64.5 min of PA, children in this sample who walked to school participated in 24.7 min higher PA during the day. Further, overweight children who walked to school took 43.7 min higher PA during the day than overweight children who were driven to school. Discussion This study indicated that across the entire day Cypriot boys took an average of 15,221 steps per day while the girls accumulated 11,341 steps per day. These levels are lower than previous levels reported among New Zealand (Duncan et al., 2006) and Sweden, but are comparable to levels reported in Australia and higher than those reported in the US (Vincent et al., 2003). Approximately three quarters of boys and half of girls met the PCPFS criteria. Although our data were collected in winter and previous studies indicate that step counts are lower in winter (Loucaides et al., 2003), the current findings suggest that activity levels are lower than those recorded in many other countries with marked gender differences. For the after school and the whole day activity, there were significant interactions between weight status and travel mode to school. Differences in these two periods between children who walked to school and traveled to school by car were evident only for overweight children. Overweight children who
walked to school were more active during the after school hours and during the whole day in comparison to overweight children who were driven to school. Thus, promoting walking to school may be an effective means of increasing physical activity among overweight children. This is the first study to examine the effect of weight status on travel mode to school and evidence from other studies is needed to further explore this relationship. It should also be noted that mean BMI in our study was higher than mean BMI in other studies examining this relationship (Tudor-Locke et al., 2003; Cooper et al., 2003 2005) and the prevalence of overweight in this sample was higher than national estimates (Savva et al., 2002). It is also important to acknowledge that only a quarter of children in this sample of Cypriot children reported walking to school in comparison to 40.0%, 64.0%, 59.0% of Filipino, English and Danish children (Tudor-Locke et al., 2003; Cooper et al., 2003, 2005). Therefore, in light of the differences in step counts between walkers and non-walkers, encouraging more children to walk to school within the Cypriot context may be one component of intervention programs aiming for the promotion of PA among this population. Further, no children reported traveling to school by bicycle. As bicycling to school has been shown to have a positive effect on children's cardiovascular fitness (Cooper et al., 2006), promoting this type of travel mode to school could have an important preventive effect on overweight and obesity. Boys took significantly more steps than girls throughout all the segments of the day. The largest differences in steps between boys and girls were observed during the 20-min recess period. Similar findings have also been reported in previous UK and US studies (Beighle et al., 2006; Tudor-Locke et al., 2006; Ridgers et al., 2005). Dividing the time allocated to recess by the number of steps recorded, boys in this study took 72 steps per minute and girls 52 steps per minute. This finding suggests that school recess periods are perhaps likely to be useful intervention locations which is consistent with previous research (Stratton and Mullan, 2005; Verstraete et al., 2006). Despite the potential importance of recess PA, it is also important to recognize that recess activity only accounted for around 9.0% of the total steps with the bulk (52.0%) being obtained after school. Thus, whereas recess period is an important segment throughout the day for promoting PA, after school time may even be more important as this accounted for half of this sample's daily steps. Study limitations and strengths Although this study has advanced knowledge by reporting activity levels, and differences by BMI group, gender and travel mode to school the study has some limitations. First, whereas pedometers are sensitive to walking behaviors or ambulatory activity they underestimate PA information involving activities such as swimming or bicycling (Bassett, 2000). In addition, pedometers do not provide information about the intensity, frequency and durations of PA (Trost, 2001; Bjornson and Belza, 2004). Children recording their own steps may have also decreased the objectivity of the PA measure, as it is impossible to certify that participants had worn the pedometers for the entire
C.A. Loucaides, R. Jago / Preventive Medicine 47 (2008) 107–111
day. Second, the distance covered when walking to school was not assessed. Third, data collection took place during winter in a single geographical location among grade five and grade six children. It is also important to highlight that participants were recruited from three schools and as such it is possible that students within a school could share similar characteristics. However, the small number of schools prevented the use of a hierarchical model in which participants are nested within schools, as such an approach would yield only three clusters. More data from other age ranges, using a larger sample, clustered in schools in different areas of the island and during different seasons would be required to generalize our findings. Conclusions This study presented descriptive data on children's steps during the segmented school day indicating that boys are more active than girls throughout all segments of the day and suggesting that PA during the different parts of the day contributes equally to both boys' and girls' daily PA. Finally, this study supports existing evidence indicating that walking to school is associated with higher daily PA levels by presenting data from a different population and using a different measure of PA behavior. Promoting active transport to school may be an important component of potential intervention programs for increasing PA but more attention is needed to examine the potential impact of child weight on the intervention effect. Acknowledgments Special thanks are due to the children and teachers who participated in this study and to Ioannis Charalambous, Lambros Stephanou and Michalis Stylianou for their help with data collection and data processing. This research was conducted as part of an elementary school research competition ‘MERA’, supported by the Cyprus Research Promotion Foundation. References Bassett, D., 2000. Validity and reliability issues in objective monitoring of physical activity. Res. Q. Exerc. Sport 71, 30–36. Beighle, A., Pangrazi, R.P., 2006. Measuring children's activity levels: the association between step counts and activity time. J. Phys. Act. Health 3, 221–229. Beighle, A., Morgan, C.F., Masurier, G.L., Pangrazi, R.P., 2006. Children's physical activity during recess and outside of school. J. Sch. Health 76, 516–520. Bjornson, K.F., Belza, B., 2004. Ambulatory activity monitoring in youth: state of the science. Pediatr. Phys. Ther. 16, 82–89. Cohen, J., 1988. Statistical power analysis for the behavioural sciences, 2nd edition. Academic Press, New York. Cole, T.J., Bellizzi, M.C., Flegal, K.M., Dietz, W.H., 2000. Establishing a standard definition for child overweight and obesity worldwide: international survey. BMJ 320, 1240–1243.
111
Cooper, A.R., Page, A.S., Foster, L.J., Qahwaji, D., 2003. Commuting to school. Are children who walk more physically active? Am. J. Prev. Med. 25, 273–276. Cooper, A.R., Andersen, L.B., Wedderkopp, N., Page, A.S., Froberg, K., 2005. Physical activity levels of children who walk, cycle, or are driven to school. Am. J. Prev. Med. 29, 179–184. Cooper, A.R., Wedderkopp, N., Wang, H., Andersen, L.B., Froberg, K., Page, A.S., 2006. Active travel to school and cardiovascular fitness in Danish children and adolescents. Med. Sci. Sports Exerc. 38, 1724–1731. Duncan, J.S., Schofield, G., Duncan, E.K., 2006. Pedometer-determined physical activity and body composition in New Zealand children. Med. Sci. Sports Exerc. 38, 1402–1409. Eston, R., Rowlands, A., Ingledew, D., 1998. Validity of heart rate, pedometry, and accelerometry for predicting the energy cost of children's activities. J. Appl. Physiol. 84, 362–371. Jago, R., Baranowski, T., 2004. Non-curricular approaches for increasing physical activity in youth: a review. Prev. Med. 39, 157–163. Loucaides, C.A., Chedzoy, S., Bennett, N., 2003. Pedometer-assessed physical (ambulatory) activity in Cypriot children. Eur. Phy. Educ. Rev. 9, 43–55. President's Council on Physical Fitness and Sports, 2001. The President's Challenge Physical Activity and Fitness Awards Program. President's Council on Physical Fitness and Sports, U.S. Department of Health and Human Services. Bloomington, IN. Ridgers, D.N., Stratton, G., Fairclough, S.J., 2005. Assessing physical activity during recess using accelerometery. Prev. Med. 41, 102–107. Rowe, D.A., Mahar, M.T., Raedeke, T.D., Lore, J., 2004. Measuring physical activity in children with pedometers. Reliability, reactivity, and replacement of missing data. Pediatr. Exerc. Sci. 16, 1–12. Savva, S.C., Kourides, Y., Tornaritis, M., Epiphaniou-Savva, M., Chadjigeorgiou, C., Kafatos, A., 2002. Obesity in children and adolescents in Cyprus. Prevalence and predisposing factors. Int. J. Obes. 26, 1036–1045. Schneider, P.L., Crouter, S.E., Lukajic, O., Bassett, D.R., 2003. Accuracy and reliability of 10 pedometers for measuring steps over a 400-m walk. Med. Sci. Sports Exerc. 35, 1779–1784. Stratton, G., Mullan, E., 2005. The effect of multicolour markings on children's physical activity level during recess. Prev. Med. 41, 828–833. Strong, W.B., Malina, R.M., Blimkie, C.J.R., et al., 2005. Evidence based physical activity for school-age youth. J. Pediatr. 146, 732–737. Trost, S.G., 2001. Objective measurement of physical activity in youth: current issues, future directions. Exerc. Sport Sci. Rev. 29, 32–36. Trost, S.G., Pate, R.R., Freedson, P.S., Sallis, J.F., Taylor, W.C., 2000. Using objective physical activity measures with youth: how many days of monitoring are needed? Med. Sci. Sports Exerc. 32, 426–431. Tudor-Locke, C., 2002. Taking steps towards increased physical activity: Using pedometers to measure and motivate. President's Council on Physical Fitness and Sports Research Digest, vol. 3, pp. 1–8. Tudor-Locke, C., Ainsworth, B.E., Popkin, B.M., 2001. Active commuting to school. An overlooked source of children's physical activity? Sports Med. 31, 309–313. Tudor-Locke, C., Ainsworth, B.E., Adair, L.S., Popkin, B.M., 2003. Objective physical activity of Filipino youth stratified for commuting mode to school. Med. Sci. Sports Exerc. 35, 465–471. Tudor-Locke, C., Lee, S.M., Morgan, C.F., Beighle, A., Pangrazi, R.P., 2006. Children's pedometer-determined physical activity during the segmented school day. Med. Sci. Sports Exerc. 38, 1732–1738. Verstraete, S.J.M., Cardon, G.M., De Clercq, D.L.R., De Bourdeaudhuij, I.M.M., 2006. Increasing children's physical activity levels during recess periods in elementary schools: the effects of providing game equipment. Eur. J. Public Health 16, 415–419. Vincent, S.D., Pangrazi, R.P., Raustorp, A., Michaud Tomson, L., Cuddihy, T.F., 2003. Activity levels and body mass index of children in the United States, Sweden, and Australia. Med. Sci. Sports Exerc. 35, 1367–1373.
Preventive Medicine 47 (2008) 107 – 111 www.elsevier.com/locate/ypmed
Differences in physical activity by gender, weight status and travel mode to school in Cypriot children Constantinos A. Loucaides a,⁎, Russell Jago b a
b
Research and Evaluation Unit, Cyprus Pedagogical Institute, Nicosia, Cyprus Department of Exercise, Nutrition and Health Sciences, University of Bristol, Bristol, U.K. Available online 14 February 2008
Abstract Objective. More information about children's physical activity during different periods of the day is needed. The purpose of this study was to describe children's physical activity during the segmented school day and examine potential differences during different periods of the day across gender, travel mode to school and weight status. Methods. School children (N = 247) wore pedometers for four consecutive school days and recorded their steps during the before school period, the 20-min school break, the whole school period, the after school period and for the whole day. Children also reported how they traveled to school. Data were collected in Cyprus, in January/February of 2007. Results. T-tests indicated that boys took significantly higher steps than girls throughout all the segments of the day ( p b 0.001) and children who walked to school exhibited higher step counts during the before school period ( p b 0.001), the after school period ( p b 0.01), and the whole day ( p b 0.01) in comparison to children who used motorized transport. A three-way ANOVA revealed significant two-way interactions between weight status and travel mode to school. Conclusion. Results suggest that there is a need to promote physical activity among children, especially among girls. Promoting active transport to school may further enhance the effectiveness of intervention programs. © 2008 Elsevier Inc. All rights reserved. Keywords: Pedometer; Children; Travel mode to school; Gender; Weight status
Introduction Participation in physical activity (PA) can promote children's musculoskeletal health, several components of cardiovascular health, reduce adiposity in overweight youth and improve blood pressure in mildly hypertensive adolescents (Strong et al., 2005). Promoting PA behavior among children in Cyprus could help reduce increased levels of overweight and obesity (Savva et al., 2002). Monitoring children's PA is an important first step for promoting this behavior. An objective and cost effective option for assessing children's PA is the pedometer. Electronic pedometers provide valid assessments of volume of PA performed by children (Trost, 2001; Bjornson and Belza, 2004). Identifying which periods of the day children are the least active could help target these ⁎ Corresponding author. 77 Larnaca Avenue, Aglanjia, 2102, Nicosia, Cyprus. E-mail address: [email protected] (C.A. Loucaides). 0091-7435/$ - see front matter © 2008 Elsevier Inc. All rights reserved. doi:10.1016/j.ypmed.2008.01.025
periods for promoting PA behavior. A recent study has quantified children's PA levels during distinct periods throughout the school day including the before school period, recess, lunch, physical education and after school activity (TudorLocke et al., 2006), but the study is limited by a small sample obtained from a single elementary school. No study has examined children's pedometer-determined PA during the segmented school day in relation to travel mode to school. Because active commuting to school is an important source of children's daily PA (Tudor-Locke et al., 2001; Jago and Baranowski, 2004), it would be interesting to examine whether children who walk to school differ in steps taken during specific periods of the day in comparison to those children who are driven to school. Previous studies using accelerometers in British (Cooper et al., 2003) and Danish children (Cooper et al., 2005), have shown that boys and girls who walked to school exhibited higher levels of overall PA in comparison to those with motorized transport. Providing data on travel mode to
108
C.A. Loucaides, R. Jago / Preventive Medicine 47 (2008) 107–111
school and PA from a different cultural perspective may help strengthen existing evidence. The aims of this study were to: 1) describe boys' and girls' pedometer-determined PA during the segmented school day, and 2) examine possible differences in PA levels during the segmented school day across gender, weight status and travel mode to school. Methods Participants Grade five (mean age 10.6 ± 0.3) and grade six (mean age 11.6 ± 0.3) Greek– Cypriot children from three urban public elementary schools in the coastal town of Lemesos, Cyprus were invited to participate in this study. In total, 296 children returned signed informed consent forms. Participation rates from the three schools were 75.2%, 73.9% and 66.4%. The Cyprus Pedagogical Institute and the Cyprus Research Promotion Foundation approved the protocol for this study.
Procedure PA was measured using an electronic pedometer (DW-200, Yamax Corporation, Tokyo, Japan). This pedometer has been shown to be valid and reliable among both children (Eston et al., 1998) and adults (Schneider et al., 2003). Prior to use, all pedometers were checked for accuracy and exhibited less than 5% error on a walk test and less than 1% error on a shake test (Tudor-Locke, 2002). Children were requested to wear the pedometers for four consecutive days. This monitoring frame has been shown to provide reliable estimates of children's PA (Trost et al., 2000). Data recording took place from Monday morning through to Thursday evening. In order to familiarize the children with the instrument, pedometers were given to the children the Friday of the previous week and were asked to fit the pedometer on Monday after awakening and dressing. Children were instructed to fit the pedometer at the waist and make sure that the cover was closed. They were also instructed to remove the pedometers only when bathing, swimming and sleeping. Instructions were given on how to reset the pedometer before reattaching it each morning. Children were given a recording card and were asked to record their steps during specific times of the day. The first recording was at 07:45 just before the school day started. This value represented children's before school activity. The second recording took place at 09:05, just before the 20-min break and the third recording at 09:25, just after the end of the break. The difference between these two values represented recess activity. The fourth recording took place at 13:05 just before the end of the school day. The difference between this value and the value recorded by children at 07:45, represented school activity. The fifth recording took place just before children went to bed. This value represented whole day activity and the difference between school activity and whole day activity represented after school activity. During school time, the recording cards were kept in the classrooms and class teachers supervised recording compliance. Two more 10-min breaks (10:45–10:55 and 12:15–12:25) are offered to
children during the school day. Children also noted how they traveled to school each morning (walk or car) and a further question also asked children to note their usual mode of travel to school (walk or car), in order to examine children's consistency in their reports. Height and weight were measured using a portable stadiometer and scale (SECA - Model 713) and Body Mass index (BMI) was calculated (kg/m2).
Statistical analyses Participants were included in the analyses if they had complete data for the 4 days of measurement. After screening the pedometer recording sheets, 38 children had incomplete data. Missing data were due to children being absent from school, children forgetting to fit the pedometer in the morning or losing the pedometer. Seven children were excluded from the analyses because of unreasonable step recordings (e.g., recording less step counts in the after break recording in comparison to the before break recording). Four participants were also excluded because of extremely high (N35,000) step counts per day (Rowe et al., 2004). Complete data were available for 247 (83.4%) children (124 boys and 123 girls). The percentage of boys and girls achieving the thresholds proposed by the President's Council on Physical Fitness and Sport (2001) (PCPFS) of 11,000 steps per day for girls and 13,000 steps per day for boys were calculated. BMI was computed as kg/m2. Children were classified as overweight and obese based on cut off points for BMI developed by the International Obesity Task Force (Cole et al., 2000). Means and standard deviations were computed for step counts across all five periods of the day. Independent-sample t-tests were conducted to examine potential differences in mean steps between boys and girls and between children who walked to school and those who were driven to school across the five periods. Finally, three-way ANOVAs were conducted to examine possible differences in steps in each of the five periods of the day across gender, travel mode to school and weight status. These were 2 (boys, girls) × 2 (car, walk) × 2 (normal weight, overweight) between subjects ANOVAs. Effect sizes (η2) were also calculated to examine the practical significance of the differences between group means (Cohen, 1988).
Results Mean BMI was 19.8 ± 3.8. The proportion of obese children in this sample was identical to national estimates (6.8 % versus 6.7%) but the proportion of overweight children in this sample was higher (29.6% versus 20.8%) (Savva et al., 2002). Sixty five of the children or 26.3 % reported traveling to school by foot and 179 or 72.5% reported traveling to school by car. About three quarters of boys (73.4%) exceeded the PCPFS threshold of 13,000 steps per day and 49.6% of girls exceeded the PCPFS threshold of 11,000 steps per day. Table 1 presents means, standard deviations and effect sizes of steps taken throughout the five different periods across boys and girls and the whole sample. Boys took significantly more
Table 1 Descriptives of steps and effect sizes during the segmented day across boys and girls and the whole sample in Cypriot children in January/February of 2007
Before school (awakening–07:45) 20-min Recess (09:05–09:25) School (07:45–13:05) After school (13:05–bed time) Whole day (awakening–bed time)
Boys (n = 124)
Girls (n = 123)
Mean (SD)
Mean (SD)
1227 (610) 1445 (585) 6030 (1873) 7964 (3353) 15221 (4468)
947 (423)⁎⁎⁎ 1039 (422)⁎⁎⁎ 4450 (1314)⁎⁎⁎ 5944 (2574)⁎⁎⁎ 11341 (3192)⁎⁎⁎
Effect Size (η2)a
All (n = 247) Mean (SD)
0.07 0.14 0.19 0.10 0.20
1088 (542) 1243 (549) 5243 (1799) 6958 (3151) 13289 (4338)
⁎⁎⁎Significant difference at the p b 0.001 level between boys and girls. Effect sizes of the significant differences in means between boys and girls. Values of 0.01, 0.06, and 0.14 can be interpreted as small, medium, and large effect sizes, respectively.
a
C.A. Loucaides, R. Jago / Preventive Medicine 47 (2008) 107–111 Table 2 Descriptives of steps and effect sizes during the segmented day across children who were driven and those who walked to school in Cypriot children in January/ February of 2007 Travel by car (n = 179) a
Walk (n = 65)
Mean (SD)
Mean (SD)
Before school 983 (487) 1355 (597)⁎⁎⁎ (awakening–07:45) 20-min Recess (09:05–09:25) 1244 (539) 1231 (584) School (07:45–13:05) 5215 (1861) 5329 (1653) After school (13:05–bed time) 6596 (2940) 8026 (3530)⁎⁎ Whole day 12795 (4290) 14710 (4278)⁎⁎ (awakening–bed time)
Effect Size (η2) b
0.08 – – 0.03 0.04
⁎⁎Significant difference at the p b 0.01 level between children who were driven and children who walked to school. ⁎⁎⁎Significant difference at the p b 0.001 level between children who were driven and children who walked to school. a Three children were inconsistent in their reports of travel mode to school and were thus excluded from the analysis. b Effect sizes of the significant differences in means between children who were driven and children who walked to school. Values of 0.01, 0.06, and 0.14 can be interpreted as small, medium, and large effect sizes, respectively.
steps than girls throughout all the periods of the day ( p b 0.001). Before school activity contributed to about 8.0% of both boys' and girls' daily PA, the 20-min recess period contributed to about 9.0% of daily PA, the rest of the school day about 31.0% of total daily activity and the after school activity about 52.0% of total daily activity. Table 2 presents means, standard deviations and effect sizes of steps taken throughout the five different periods across children who walked to school and children who were driven to school. Children who walked to school took significantly more steps than children who were driven to school during the before school ( p b 0.001), the after school ( p b 0.01), and the whole day period ( p b 0.01). Results of the three-way ANOVA (see Table 3) for the before school steps revealed a significant main effect for gender ( p b 0.001, η2 = 0.06), and a significant main effect for travel mode to school ( p b 0.001, η2 = 0.07). Boys took significantly more steps than girls and children who walked to school took significantly more steps than children who were driven to school. Analyses for the 20-min recess steps revealed a significant main effect for gender ( p b 0.001, η2 = 0.14). Identical results were revealed for the school steps ( p b 0.001, η2 = 0.16), with boys taking more steps than girls. The three-way ANOVA for the after school steps revealed a significant main effect for gender, ( p b 0.001, η2 = 0.10), a significant main effect for travel mode to school, ( p b 0.01, η2 = 0.04) and a significant travel mode by weight status interaction effect (see Table 3). Identical results were revealed for the whole day steps analysis. The ANOVA yielded a significant main effect for gender, ( p b 0.001, η2 = 0.19), a significant main effect for travel mode to school, ( p b 0.01, η2 = 0.04) and a significant travel mode by weight status interaction effect. Boys took more steps than girls and those who walked to school took more steps than those who were driven to school. The interactions showed that children who walked to school gained more steps than those who were driven during the after school
109
Table 3 Descriptives of steps during the segmented day across gender, travel mode and weight status and F-values of interaction effects in Cypriot children in January/ February of 2007 Boys Mean (SD)
Girls Mean (SD)
Normal weight
Overweight
Normal weight
Overweight
Before school Car 1148 (550)
1028 (589)
887 (390)
868 (396)
Walk 1472 (454)
1577 (932)
1158 (399)
1069 (567)
20-min Recess Car 1456 (596)
1394 (624)
1124 (453)
970 (376)
Walk 1495 (681)
1456 (326)
809 (196)
925 (523)
F-values (interaction effects)
F(1,226) = 1.59, p = 0.21 a F(1,226) = 0.09, p = 0.77 b F(1,226) = 0.23, p = 0.63 c F(1,226) = 0.86, p = 0.36 d
F(1,226) = 2.01, p = 0.16 a F(1,226) = 0.04, p = 0.85 b F(1,226) = 0.81, p = 0.37 c F(1,226) = 0.58, p = 0.45 d
School Car 6152 (2030) 5736 (2132) 4604 (1367) 4143 (1084) F(1,226) = 0.20, p = 0.66 a Walk 6074 (1477) 6008 (1687) 4327 (1430) 4155 (1095) F(1,226) = 0.02, p = 0.88 b F(1,226) = 0.39, p = 0.53 c F(1,226) = 0.00, p = 0.95 d After school Car 8212 (3433) 6372 (2648) 6207 (2438) 4791 (1977) F(1,226) = 0.95, p = 0.33 a Walk 8717 (3236) 9661 (3935) 5861 (2118) 7176 (4142) F(1,226) = 0.20, p = 0.66 b F(1,226) = 9.37, p = 0.01 c⁎⁎ F(1,226) = 0.00, p = 0.98 d Whole day Car 15512 (4769) 13136 (4319) 11699 (3117) 9802 (2686) F(1,226) = 1.24, p = 0.27 a Walk 16263 (3598) 17245 (4016) 11345 (2911) 12400 (4252) F(1,226) = 0.06, p = 0.82 b F(1,226) = 7.18, p = 0.01 c⁎⁎ F(1,226) = 0.03, p = 0.86 d ⁎⁎Significant at the p b 0.01 level (η2 = 0.04 for the after school steps and η2 = 0.03 for the whole day steps). a Gender by travel mode to school interaction effect. b Gender by weight status interaction effect. c Travel mode to school by weight status interaction effect. d Gender by travel mode to school by weight status interaction effect.
110
C.A. Loucaides, R. Jago / Preventive Medicine 47 (2008) 107–111
Fig. 1. Bar chart showing the interaction between travel mode to school and weight status during the whole day activity in Cypriot children in January/ February of 2007.
and the whole day steps, but this difference was only evident among overweight children ( p b 0.01). The interaction for the whole day steps is presented graphically in Fig. 1. Based on the regression equation by Beighle and Pangrazi (2006) suggesting that 5000 steps results in the accumulation of 64.5 min of PA, children in this sample who walked to school participated in 24.7 min higher PA during the day. Further, overweight children who walked to school took 43.7 min higher PA during the day than overweight children who were driven to school. Discussion This study indicated that across the entire day Cypriot boys took an average of 15,221 steps per day while the girls accumulated 11,341 steps per day. These levels are lower than previous levels reported among New Zealand (Duncan et al., 2006) and Sweden, but are comparable to levels reported in Australia and higher than those reported in the US (Vincent et al., 2003). Approximately three quarters of boys and half of girls met the PCPFS criteria. Although our data were collected in winter and previous studies indicate that step counts are lower in winter (Loucaides et al., 2003), the current findings suggest that activity levels are lower than those recorded in many other countries with marked gender differences. For the after school and the whole day activity, there were significant interactions between weight status and travel mode to school. Differences in these two periods between children who walked to school and traveled to school by car were evident only for overweight children. Overweight children who
walked to school were more active during the after school hours and during the whole day in comparison to overweight children who were driven to school. Thus, promoting walking to school may be an effective means of increasing physical activity among overweight children. This is the first study to examine the effect of weight status on travel mode to school and evidence from other studies is needed to further explore this relationship. It should also be noted that mean BMI in our study was higher than mean BMI in other studies examining this relationship (Tudor-Locke et al., 2003; Cooper et al., 2003 2005) and the prevalence of overweight in this sample was higher than national estimates (Savva et al., 2002). It is also important to acknowledge that only a quarter of children in this sample of Cypriot children reported walking to school in comparison to 40.0%, 64.0%, 59.0% of Filipino, English and Danish children (Tudor-Locke et al., 2003; Cooper et al., 2003, 2005). Therefore, in light of the differences in step counts between walkers and non-walkers, encouraging more children to walk to school within the Cypriot context may be one component of intervention programs aiming for the promotion of PA among this population. Further, no children reported traveling to school by bicycle. As bicycling to school has been shown to have a positive effect on children's cardiovascular fitness (Cooper et al., 2006), promoting this type of travel mode to school could have an important preventive effect on overweight and obesity. Boys took significantly more steps than girls throughout all the segments of the day. The largest differences in steps between boys and girls were observed during the 20-min recess period. Similar findings have also been reported in previous UK and US studies (Beighle et al., 2006; Tudor-Locke et al., 2006; Ridgers et al., 2005). Dividing the time allocated to recess by the number of steps recorded, boys in this study took 72 steps per minute and girls 52 steps per minute. This finding suggests that school recess periods are perhaps likely to be useful intervention locations which is consistent with previous research (Stratton and Mullan, 2005; Verstraete et al., 2006). Despite the potential importance of recess PA, it is also important to recognize that recess activity only accounted for around 9.0% of the total steps with the bulk (52.0%) being obtained after school. Thus, whereas recess period is an important segment throughout the day for promoting PA, after school time may even be more important as this accounted for half of this sample's daily steps. Study limitations and strengths Although this study has advanced knowledge by reporting activity levels, and differences by BMI group, gender and travel mode to school the study has some limitations. First, whereas pedometers are sensitive to walking behaviors or ambulatory activity they underestimate PA information involving activities such as swimming or bicycling (Bassett, 2000). In addition, pedometers do not provide information about the intensity, frequency and durations of PA (Trost, 2001; Bjornson and Belza, 2004). Children recording their own steps may have also decreased the objectivity of the PA measure, as it is impossible to certify that participants had worn the pedometers for the entire
C.A. Loucaides, R. Jago / Preventive Medicine 47 (2008) 107–111
day. Second, the distance covered when walking to school was not assessed. Third, data collection took place during winter in a single geographical location among grade five and grade six children. It is also important to highlight that participants were recruited from three schools and as such it is possible that students within a school could share similar characteristics. However, the small number of schools prevented the use of a hierarchical model in which participants are nested within schools, as such an approach would yield only three clusters. More data from other age ranges, using a larger sample, clustered in schools in different areas of the island and during different seasons would be required to generalize our findings. Conclusions This study presented descriptive data on children's steps during the segmented school day indicating that boys are more active than girls throughout all segments of the day and suggesting that PA during the different parts of the day contributes equally to both boys' and girls' daily PA. Finally, this study supports existing evidence indicating that walking to school is associated with higher daily PA levels by presenting data from a different population and using a different measure of PA behavior. Promoting active transport to school may be an important component of potential intervention programs for increasing PA but more attention is needed to examine the potential impact of child weight on the intervention effect. Acknowledgments Special thanks are due to the children and teachers who participated in this study and to Ioannis Charalambous, Lambros Stephanou and Michalis Stylianou for their help with data collection and data processing. This research was conducted as part of an elementary school research competition ‘MERA’, supported by the Cyprus Research Promotion Foundation. References Bassett, D., 2000. Validity and reliability issues in objective monitoring of physical activity. Res. Q. Exerc. Sport 71, 30–36. Beighle, A., Pangrazi, R.P., 2006. Measuring children's activity levels: the association between step counts and activity time. J. Phys. Act. Health 3, 221–229. Beighle, A., Morgan, C.F., Masurier, G.L., Pangrazi, R.P., 2006. Children's physical activity during recess and outside of school. J. Sch. Health 76, 516–520. Bjornson, K.F., Belza, B., 2004. Ambulatory activity monitoring in youth: state of the science. Pediatr. Phys. Ther. 16, 82–89. Cohen, J., 1988. Statistical power analysis for the behavioural sciences, 2nd edition. Academic Press, New York. Cole, T.J., Bellizzi, M.C., Flegal, K.M., Dietz, W.H., 2000. Establishing a standard definition for child overweight and obesity worldwide: international survey. BMJ 320, 1240–1243.
111
Cooper, A.R., Page, A.S., Foster, L.J., Qahwaji, D., 2003. Commuting to school. Are children who walk more physically active? Am. J. Prev. Med. 25, 273–276. Cooper, A.R., Andersen, L.B., Wedderkopp, N., Page, A.S., Froberg, K., 2005. Physical activity levels of children who walk, cycle, or are driven to school. Am. J. Prev. Med. 29, 179–184. Cooper, A.R., Wedderkopp, N., Wang, H., Andersen, L.B., Froberg, K., Page, A.S., 2006. Active travel to school and cardiovascular fitness in Danish children and adolescents. Med. Sci. Sports Exerc. 38, 1724–1731. Duncan, J.S., Schofield, G., Duncan, E.K., 2006. Pedometer-determined physical activity and body composition in New Zealand children. Med. Sci. Sports Exerc. 38, 1402–1409. Eston, R., Rowlands, A., Ingledew, D., 1998. Validity of heart rate, pedometry, and accelerometry for predicting the energy cost of children's activities. J. Appl. Physiol. 84, 362–371. Jago, R., Baranowski, T., 2004. Non-curricular approaches for increasing physical activity in youth: a review. Prev. Med. 39, 157–163. Loucaides, C.A., Chedzoy, S., Bennett, N., 2003. Pedometer-assessed physical (ambulatory) activity in Cypriot children. Eur. Phy. Educ. Rev. 9, 43–55. President's Council on Physical Fitness and Sports, 2001. The President's Challenge Physical Activity and Fitness Awards Program. President's Council on Physical Fitness and Sports, U.S. Department of Health and Human Services. Bloomington, IN. Ridgers, D.N., Stratton, G., Fairclough, S.J., 2005. Assessing physical activity during recess using accelerometery. Prev. Med. 41, 102–107. Rowe, D.A., Mahar, M.T., Raedeke, T.D., Lore, J., 2004. Measuring physical activity in children with pedometers. Reliability, reactivity, and replacement of missing data. Pediatr. Exerc. Sci. 16, 1–12. Savva, S.C., Kourides, Y., Tornaritis, M., Epiphaniou-Savva, M., Chadjigeorgiou, C., Kafatos, A., 2002. Obesity in children and adolescents in Cyprus. Prevalence and predisposing factors. Int. J. Obes. 26, 1036–1045. Schneider, P.L., Crouter, S.E., Lukajic, O., Bassett, D.R., 2003. Accuracy and reliability of 10 pedometers for measuring steps over a 400-m walk. Med. Sci. Sports Exerc. 35, 1779–1784. Stratton, G., Mullan, E., 2005. The effect of multicolour markings on children's physical activity level during recess. Prev. Med. 41, 828–833. Strong, W.B., Malina, R.M., Blimkie, C.J.R., et al., 2005. Evidence based physical activity for school-age youth. J. Pediatr. 146, 732–737. Trost, S.G., 2001. Objective measurement of physical activity in youth: current issues, future directions. Exerc. Sport Sci. Rev. 29, 32–36. Trost, S.G., Pate, R.R., Freedson, P.S., Sallis, J.F., Taylor, W.C., 2000. Using objective physical activity measures with youth: how many days of monitoring are needed? Med. Sci. Sports Exerc. 32, 426–431. Tudor-Locke, C., 2002. Taking steps towards increased physical activity: Using pedometers to measure and motivate. President's Council on Physical Fitness and Sports Research Digest, vol. 3, pp. 1–8. Tudor-Locke, C., Ainsworth, B.E., Popkin, B.M., 2001. Active commuting to school. An overlooked source of children's physical activity? Sports Med. 31, 309–313. Tudor-Locke, C., Ainsworth, B.E., Adair, L.S., Popkin, B.M., 2003. Objective physical activity of Filipino youth stratified for commuting mode to school. Med. Sci. Sports Exerc. 35, 465–471. Tudor-Locke, C., Lee, S.M., Morgan, C.F., Beighle, A., Pangrazi, R.P., 2006. Children's pedometer-determined physical activity during the segmented school day. Med. Sci. Sports Exerc. 38, 1732–1738. Verstraete, S.J.M., Cardon, G.M., De Clercq, D.L.R., De Bourdeaudhuij, I.M.M., 2006. Increasing children's physical activity levels during recess periods in elementary schools: the effects of providing game equipment. Eur. J. Public Health 16, 415–419. Vincent, S.D., Pangrazi, R.P., Raustorp, A., Michaud Tomson, L., Cuddihy, T.F., 2003. Activity levels and body mass index of children in the United States, Sweden, and Australia. Med. Sci. Sports Exerc. 35, 1367–1373.